My professor teaches his medical registrars and students that you should
aim to achieve an 80% response rate or over in a survey.
This is the target and is part of the research design. You return to your
non-responders at designated intervals with ?planned inducements, or use
another route ie telephone/email. This will be more than once and is
planned for.
This strategy should be documented to show that all necessary steps were
taken to guard against a poor response. A large group of non-responders may
infer a problem with the research instrument or that there is a group in
your original sample which are profoundly different in some way. A low
response produces doubt in the experienced reader as to the validity of the
results.
I do know some doctors who will dismiss any survey if the response rate
falls below their own level of 'tolerance'.
If they aim high i don't see why we cannot!
best wishes
Katharine Wylie
http://www.bestbets.org
|